Projects

Built an interactive clustering visualizer which implements k-means and DBSCAN (density-based spatial clustering of applications with noise). Used paper.js for canvas graphics.

Implemented a random forest and XGBoost model to predict housing prices using the Ames Housing Dataset. Completed data cleaning, feature engineering, model tuning to achieve top 10% of submitted work on Kaggle competition.

Designed and breadboarded an EOG (electrooculography) acquisition circuit as part of my independent study in Spring 2022. The circuit may be used to track gaze through a series of electrodes around subject's eyes. This design includes a driven right leg circuit in order to reduce the high common-mode interference of the EOG signals.

Built a series of maximum likelihood models to classify spoken Arabic Digits using Cepstral Coefficients (see slides in GitHub repository). Achieved 84% accuracy on test set using k-means maximum likelihood classifier, and 96% using a Gaussian Mixture Model. Submitted as a class project for ECE 480, Applied Probability for Statistical Learning.

Made a pong game with few additional features using JS, HTML Canvas, and CSS. See https://github.com/evanglas/Pong for repository.

My team and I at HackDuke21 built a website designed to connect instrument donors with underprivledged students. I worked on the frontend was built with ReactJS, HTML, and CSS. See https://moremusic.vercel.app/ for running version.

This is a slide deck I presented Spring semester 2021 for a club's onboarding process. I used this deck to present the results of my data analysis on a mock data set of customer histories for a credit card company.

Click to see the code for this website! At the moment, the website was built using HMTL, CSS, and a bit of JavaScript.

This is a portion of the final deliverable for my Spring 2021 DIIG (Duke Impact Investing Group) Project. That semester, I conducted data analysis for a housing startup along with a small team. These slides consist of my portion of the originally larger slide deck, and were presented to individuals within the startup. I completed all analysis shown in the slides.

This is the result of my final project for my Introduction to Engineering Class. I worked in a team of four to solve a design problem given to us by an outside client. Our project was to design a soiling sensor, which detected the loss of efficiency of solar panels due to dust/accumulated debris.

This is a song I made in MATLAB for an ECE 280 (Signals and Systems) class project. I recreated the first 40 seconds of "The Swan" using only elementary signals and some signal processing